Spaces:
Running
on
T4
Running
on
T4
Tom Aarsen
commited on
Commit
·
810f572
1
Parent(s):
ebb05ef
Wrap details/summary in HTML instead
Browse files
app.py
CHANGED
@@ -645,7 +645,11 @@ with gr.Blocks(
|
|
645 |
|
646 |
Sentence Transformers embedding models can be optimized for **faster inference** on CPU and GPU devices by exporting, quantizing, and optimizing them in ONNX and OpenVINO formats.
|
647 |
Observe the [Speeding up Inference](https://sbert.net/docs/sentence_transformer/usage/efficiency.html) documentation for more information.
|
648 |
-
|
|
|
|
|
|
|
|
|
649 |
<details><summary>Click to see performance benchmarks</summary>
|
650 |
|
651 |
<table>
|
@@ -667,18 +671,16 @@ Observe the [Speeding up Inference](https://sbert.net/docs/sentence_transformer/
|
|
667 |
</tbody>
|
668 |
</table>
|
669 |
|
670 |
-
|
671 |
-
|
672 |
-
|
673 |
-
|
674 |
-
|
|
|
|
|
675 |
|
676 |
</details>
|
677 |
-
|
678 |
-
""",
|
679 |
-
label="",
|
680 |
-
container=True,
|
681 |
-
)
|
682 |
|
683 |
model_id = HuggingfaceHubSearch(
|
684 |
label="Hub Model ID",
|
|
|
645 |
|
646 |
Sentence Transformers embedding models can be optimized for **faster inference** on CPU and GPU devices by exporting, quantizing, and optimizing them in ONNX and OpenVINO formats.
|
647 |
Observe the [Speeding up Inference](https://sbert.net/docs/sentence_transformer/usage/efficiency.html) documentation for more information.
|
648 |
+
""",
|
649 |
+
label="",
|
650 |
+
container=True,
|
651 |
+
)
|
652 |
+
gr.HTML(value="""\
|
653 |
<details><summary>Click to see performance benchmarks</summary>
|
654 |
|
655 |
<table>
|
|
|
671 |
</tbody>
|
672 |
</table>
|
673 |
|
674 |
+
<ul>
|
675 |
+
<li><code>onnx</code> refers to the ONNX backend</li>
|
676 |
+
<li><code>onnx-qint8</code> refers to ONNX (Dynamic Quantization)</li>
|
677 |
+
<li><code>onnx-O1</code> to <code>onnx-O4</code> refers to ONNX (Optimization)</li>
|
678 |
+
<li><code>openvino</code> refers to the OpenVINO backend</li>
|
679 |
+
<li><code>openvino-qint8</code> refers to OpenVINO (Static Quantization)</li>
|
680 |
+
</ul>
|
681 |
|
682 |
</details>
|
683 |
+
""")
|
|
|
|
|
|
|
|
|
684 |
|
685 |
model_id = HuggingfaceHubSearch(
|
686 |
label="Hub Model ID",
|